“We currently have no plans to support Xwayland”
devtalk.nvidia.com
devtalk.nvidia.com
Intel is actively trying to kill off lowend and mid-range GPUs in the desktop, and has locked them out in the mobile space, leaving Nvidia with just the high end GPU market and their ventures into the ARM SOC space.
By not treating Linux end user devices as first class citizens, I don't see how Nvidia intends to have a viable business long term. Who wants a tablet, phone or desktop GPU without drivers?
They're a business, so when their customers need it, they'll support it. Until then, only hobbyists or an insufficient part of their customer base is interested, so understandably, they're not investing in it.
I highly doubt they'll want to stay on something that isn't seeing much in the way of support, so time for a new SOC vendor...
Moreover, if you have an NVIDIA GPU, you most likely need a decent GPU for something - high-end video games, 3D modelling work, GPU-accelerated computations, etc. Linux distros are, frankly, not much different for one another - they're pretty interchangeable for almost all tasks I can think of as a user/developer. GPUs, on the other hand, are a bit less interchangeable between vendors. You really have only two players to choose from (Intel GPUs don't count for anything serious) - so if both suddenly become incompatibile with a given Linux distro, then it's obviously bye bye to that Linux distro.
Will Toyota go for running legacy, unmaintained Xorg in their cars? How about BMW? I don't see Nvidia ready to dump this market they've spent a decade building, so short of them being complete idiots, they will probably walk back on this.
That being said, these were the people that tried to pass off a painted block of wood as a GPU on multiple occasions: https://semiaccurate.com/2011/03/11/nvidias-ceo-has-another-...
Given this -- if Toyota doesn't want to run "legacy, unmaintained" X11 in their cars, then.... they don't have to, they can just move to Wayland, which will work, and NVidia supports Wayland just fine -- and they don't have to worry about X11 at all anymore?
So, everything would be fine? I'm not sure what point you're making here...
Gaming enthusiasts aren't running Linux. The HPC/GPU compute market don't care about Wayland. Android devices aren't using Wayland.
What part of their target market runs Linux?
What portion of the desktop market runs Linux?
If it wasn't for Linux, I'd be in a very different profession, so I'm a huge fan. But I don't understand where you guys that act like the Linux desktop is a big enough market that this will make a material affect on nvidia's bottom line is so weird to me.
More and more do today.
And the GP comment was right anyway, 90% of the Market is for sure not running Linux for gaming on PC. 1~2% is nice but it's hardly enough to convince large publishers to support Linux.
If you noticed, the main article speaks about how those "teams of developers" DON'T plan to support XWayland.
And what's important is sales potential, which you can evaluate by sales percentages per OS for cross platform games. Steam doesn't publish such numbers. Humble Bundle does though.
People who use their Android phones to purchase and pay Humble Bundle, get classified under "Linux". You can see big Linux platform numbers even for bundles that don't have a single Linux game.
I wouldn't be surprised if a survey actually overestimated the number of Linux users. Using Linux for gaming is a choice and usually a more difficult path than gaming on Windows. So, Linux users are likely move vocal about their use of Linux, for various reasons: (1) it's often a principled choice; (2) by showing that a significant chunk uses Linux, more gaming studios will target Linux.
tl;dr: it would be nice if Valve released actual stats, but I think the survey is a good ballpark figure. In other words, the market is pretty much non-existent.
> by showing that a significant chunk uses Linux, more gaming studios will target Linux.
That's already happening, if you paid attention.
No matter which way you slice it, the Linux gaming community is tiny when compared to Windows and consoles. Sure, it's growing, and there are more and more games coming to the OS, and more choices = better of course.
But we are a long way from Linux being a platform that matters in the overall computer gaming world.
Linux gaming is obviously a subject that you hold dear - you've jumped in multiple times on almost every single top thread in this discussion - and that's ok, but ease off a bit..
Let your points stand on their own merit, you don't need to call people out because they're not as deeply informed on the subject matter as you are, or because they don't agree with your perspective.
Just to speak against this notion (which absolutely used to be true, no contest there): anecdotally, I don't find that to be true anymore, at least using Steam.
It works really smoothly for me, just as on Windows. In fact I'm just about ready to drop Windows as a gaming plaform. Maybe some titles aren't available for Linux, but I still have enough to fall back on.
If you choose to game on Linux, you are intentionally restricting the number of games you can play. And for some people that's obviously ok, but for most (as stats show), it's not ok.
Because they need permission from the user to collect the data.
Even if the data is undercounting Linux (maybe Linux games are less likely to take the survey) the trend is consistently down for ~4 years now. So you might make a case that the size is bigger than the survey shows, but not that it is growing.
From DRM-free stores, GOG unfortunately don't publish any stats, and stats from HB show pretty good Linux percentage.
I can also play on my Windows box/console/Mac etc.
Unlikely, as the PC market is basically stable or even shrinking slowly.
I tried using Linux on several occassions over the last 10-15 years. It was never easy or fun. All the GUIs felt like they were from the 90s and nothing worked out of the box. I tried Ubunutu, Mint, Arch, and a few others. My last attempt at it was ~2014. Maybe I'll hate myself enough to try Linux again in a year or two...
I don't. I run all my games on Linux. Unlike you, I simply don't play games that don't work on Linux ;) I.e. either natively or in Wine.
You can't deny that if the big AAA players supported Linux it would bring a (large?) influx of gamers. If they'd all just target the latest Ubuntu LTS or whatever SteamOS run on, the community would make it run on their own distros.
But if you are really so desperate to spend money on those who ignore us, Wine is improving all the time, and now already approaches decent DX11 support, so even games like Overwatch are now playable in it.
However it's never been good enough. At first there just were no drivers, so I couldn't get acceptable performance from any game. Then came drivers, but I had to fight for days to get them to work on handful of games and even then performance was sub par compared to Windows. Last time drivers installed without any problems, but still from the games I tend to play only couple had native Linux support (in this case, they could be downloaded straight from Steam), but still most required Wine and that sucked the performance right out. Then there were bunch of games that didn't even work with Wine.
I'm first to admit that I do not have high end gaming PC, but 16GB of middle of the road RAM, i5 2500K, and nVidia GTX 970 serve me well on Windows side. Frankly, I see no reason to spend more money to get the same performance on Linux.
If you can get acceptable performance from your Linux PC while gaming then more power to you, but from my point of view we are still long way and I mean really long way from gaming being "viable" on Linux.
Only "native" Linux games I've played are Counter Strike: Global Offensive, Dota 2, and Insurgency. Beyond thous only game I've gotten to run smoothly enough that I could even consider playing them is Minecraft (which obviously runs even on a potato). But even with these I had some problems. CS:GO's FPS didn't quite keep above 120 all of the time and Dota 2 had few graphical glitches.
We do get our share of AAA titles, just not as many as on Windows. As it is, there are more appealing titles than I can ever play without cutting into work, chores, or (non-gaming) leisure time.
Most indie games benefit as well though; even though these do tend to work on less well-endowed GPU's, they simply look better with a heftier graphics card. I've got a desktop computer with beefy GPU exactly because I enjoy gaming; and it runs Linux, exclusively.
I also enjoyed Thimbleweed Park and I see the (excellent, btw) Wonderboy remake was released for Linux this week. Us Linux gamers are incredibly well catered for these days!
(Ubuntu // i5 2500K // GTX 970 with Intel proprietary drivers, FWIW)
-- Ron Gilbert
I don't choose a platform because it has "many games", I want to play games X, Y and Z and I want no issue, what OS should I play them with.
The answer in 99% of X, Y and Z combinations is Windows tragically
However, it's more than good enough for many games already, and Linux gaming keeps improving and growing. And not all games need super performance.
I can get around 40-50 fps in the Witcher 3 in Wine on Linux on max settings (minus hairworks), using AMD RX 480 at 1920x1200. Is that acceptable performance? It's somewhat lower than it would get on Windows on the same hardware, but it's pretty playable. More critical are rendering bugs that Wine developers are still working on.
However I've spent money on 144Hz monitors and for enough performance that the games I play run on at least constant 120 fps. Obviously it's not necessary if you are in it just for a story, but as mentioned in another post I exclusively play multiplayer games with my friends and if you haven't experienced playing games (or even just using a computer) with 144Hz (running at or higher frame rate) it's like a night and day.
I stay away from DRMed platforms, so incumbent consoles are no go for me.
About higher framerates - it should be OK, once AMD will merge FreeSync support in amdgpu driver upstream. My monitor is 60Hz anyway, but to have good experience without tearing on high frequency monitors, FreeSync can help. Today, to use it you'd need to install customized kernel from AMD.
>I stay away from DRMed platforms
I am not ideologically bound, I care about performance. If Linux is enough for you then more power to you, I wish it could be enough for me too.
If you care about performance, you'd stay away from consoles anyway. They are way too limited in comparison.
So I don't agree that Linux is much behind technically. Smaller market share is more of an issue.
As long as Linux is just sort of viable there won't be wide adoption, because why would you change everything about your computing life for something that's +/-0% gain. Even at 10% better experience you might have to consider hard if changing over is worth the effort. Linux needs to magically become superior in many ways for people to hope on board in large numbers.
It's recent, so its usage is gradually growing. Unity engine now supports it, with Unreal and Cry on the way. Once major engines will have it, it will implicitly be used in much larger numbers.
Yes, but we were talking about Linux's popularity as a gaming platform, and why it's not more widespread, and that's not an issue the huge majority of gamers share -- so they'd rather use a console.
You get near native performance in a VM, the downside being it's a little tricky to set up and you won't be able to use your 970 in Linux.
The rest are indie games tiny budget indie games.
So again, AAA or not AAA, quality is what matters, both artistic and technical.
Today, good studios are releasing games for Linux. Such as inXile, Obsidian and others.
The term AAA is analogous to Blockbuster in the film industry. It's a reference to both budget and expected unit sales. Blockbuster movies have budgets starting in the hundreds of millions where as Indie films budges are usually in the single to double digit millions. Video games are no different in that respect.
That's why I don't like the term AAA, it's way too confusing, conflating unrelated and often orthogonal issues (budget, publisher funding vs independent development, quality and so on).
Anyway, there are big budget games coming out for Linux, so above is really besides the point.
Budget is a good indicator of audience numbers regardless of quality.
You can say that there are plenty of good games on Linux and everyone should switch but the reality is that the games people want to play are either unavailable or lacking in some way that makes Linux on the desktop a non starter. C64 and Amiga probably have a larger base than Linux.
In way I don't mind that Linux gamers are getting better games on average, because of less mass market mediocrity, but more interesting (usually crowdfunded) releases that are reaching Linux gamers today. However it ties to the next point, of the market size:
> You can say that there are plenty of good games on Linux and everyone should switch but the reality is that the games people want to play are either unavailable or lacking in some way that makes Linux on the desktop a non starter.
That doesn't describe the situation well. Linux gaming is for sure a much smaller market than Windows gaming, but not only it's already viable (as many developers who sell Linux games demonstrate), it's gradually growing. So it's surely not a "non starter" anymore.
However, since it's smaller, having too many games can actually cause a problem, even if it sounds funny. Simply because it will be overcrowded developers wise, and they'll have harder time selling Linux games. So I'd say actually current amount of games healthily corresponds to the rate of Linux gaming growth.
Who says they are? There's plenty of crap available for Linux. I mentioned Braid because it's a very popular and very good Indie game available on both platforms. I surely wasn't implying that Linux only receives ports of the finest of games.
Also what makes a game "good" is subjective. From the perspective of hours played, Braid isn't very good.
> That doesn't describe the situation well. Linux gaming is for sure a much smaller market than Windows gaming, but not only it's already viable
It is viable for some small developers and for larger developer who are already using tools ported to Linux it can make sense to add support however for the majority of game studios making popular (not necessarily good or bad games) it doesn't make sense.
The market however is viable for everybody, but as I said, if everybody from the Windows gaming market will start releasing games for Linux now, it can actually hurt it, rather than improve it. Gradual growth is better, and it's already happening. Tools wise, the situation is getting much better as well, with all major engines gaining Vulkan support. Common middleware like Umbra 3D also often supports Linux these days.
Are there any numbers, I'm actually curious. I don't regularly see it on the Twitch.tv front page, so my impression was it was a bit behind the big names in esports (LoL, DotA, CS:GO, SC2, Hearthstone, Overwatch).
Here are the release dates with linux support date.
Tomb Raider: 2013, 2016
Middle-earth: Shadow of Mordor: 2014, 2015
The Talos Principle: 2014, same day
Here are some other random selection:
Borderlands 2: 2012, 2014
Europa Universalis 4: 2013, same day
Pillars of Eternity: 2015, same day
Civilization 5: 2010, 2014
Civilization 6: 2016, 2017
Factorio: not yet released, already supports both
There is a clear trend towards more games with linux support and less lag between windows release and linux release. My impression is that the tipping point for newer games is already in favour of linux support than not linux support.
If you want to argue against linux gaming, I would argue with slightly older games that didn't receive a linux port and maybe never will.
The terms gamer and gaming community are used by people to lump anyone who plays video games into a small digestible nugget. There is no single gaming community and the sheer size of each of the countless gaming communities dwarf most other hobby or enthusiast communities.
Most people promoting gaming on Linux are so out of touch that they think having Tux Racer playable makes a Linux desktop computer a first or even second class gaming platform. Having The Witcher 3 or Tomb Raider on Linux is a novelty or a way to win brownie points.
I'm not a serious gamer by any means, my highest played game only logs 403 hours of playtime, I've see people with ten times that. With that in mind, none of the games in my top 10 list in terms of hours (~1800 hours) played are available on Linux.
At this moment they make up nearly 1.3 million concurrent players. The rest of the top 10 games that do not have a linux port accumulate close to 0.7 million concurrent players.
For a lot of players, CS:GO or Dota 2 are the only games they play regularly, so linux is a viable option for them.
You'll see all kinds of posts showing Min/Max FPS and how Linux beats Windows at this game or that but you rarely see any mention of Average FPS because they're usually pretty awful.
https://www.google.com/search?q=CS%3AGO+audio+delay+linux&ie...
In any case, it's the game developer's choice and not mine. Yet another problem Linux has for developers is that there isn't always a clear choice and the flavor of the day may not be the best choice long term. That's why frameworks like Qt and Unity are so prevalent because they abstract a lot of the crap away.
I feel like a large part of the group making who believe Linux is gaming ready doesn't understand that angle.
I only play multi player games with my friends, since single player games rarely keep my interest (last single player campaign/game I finished was Portal 2 since it was so short)
Then there's always "will this even work with wine" question. Like take any Blizzard game, the downloader/launcher simply does not work on Wine (or at least it did not when I last tried it). When a new game comes out that I want to play, I don't want to think if I can play it, I want know I can if I choose to. This is even bigger concern since most interesting games now days are not AAA titles, but various smaller games, mods, and early access titles. Game we've spent most time playing last month is easily Playerunknown's Battlegrounds and with quick look at Wine DB it's status is "garbage"
If I look Linux as separate platform then there is no hope I would be using it to play games.
This is exactly the stupid mindset I'm talking about. You can play different games! It's okay, really, to play different games! Other games are fun too!
Here's an incomplete list of multiplayer games that run natively on Linux, sourced from my own steam library and the popular linux games on steam:
- Rocket League
- CS:GO (and CS:Source)
- Rust
- Garry's Mod
- Terraria
- Insurgency
- Borderlands 2 (+ pre-sequel)
- Civ V
- Left 4 Dead (1 & 2)
- Dota 2
- Torchlight II
- Team Fortress 2
- Battleblock Theater
- Natural Selection 2
- PAYDAY 2
You like multiplayer games? Great! There's loads of them on Linux, and they're all super fun. You don't need to play the same games as Windows to have fun.
To me, they are:
- license cost (a small hurdle if you're buying $60 games)
- open source (mainly a moral/ideological advantage rather than a practical one)
- developer tooling (minimized if you use tools like Docker or Vagrant)
I'll also add:
- WAY more customizable and flexible
- Can enjoy a smug sense of satisfaction as Windows users are hit by ransomware
- Does not spy on you
- Runs great on everything, including >10 year old hardware
It's also not driven by corporate interests. The #1 goal of Microsoft is to make money, and they'll put that before the needs of the user. Things like Windows 10 forced upgrades, spyware, ads in Windows explorer, etc - none of this would fly on Linux because Linux is made by people who just want the best possible operating system, not the one that monetizes uses the most effectively.
This arrogant prescriptivism is what you did last time, too. You did this last time this came up by trying to slag me because I didn't want to waste hours of my life debugging weird edge-case issues with Linux as a desktop operating system (and, funny story, Ubuntu 16.04 hard-locks on my desktop when I plug in a bog-standard USB sound card, that's cool, and has DVI timing and desync problems between my monitor and the graphics card--this is fiiiine, everything is fiiiine). You have decided it's compelling for you. That does not mean stating "it's compelling" makes it so for a single, solitary other soul.
You are frothing and should chill.
Linux isn't "my software", so why would I buy hardware for it--especially when nothing else has a problem with it? Your OS is not special no matter how much you yell at people on HN about it. Your OS doesn't matter to anyone else no matter how much you yell at people on HN about it. Attempting to put the onus of "due diligence" on the people you want to do something is just straight-up nuts--and, tbh, the same rhetorical evasion that's come out of the Linux zealot camp for decades now because it's easier than acknowledging problems and that the things you like aren't universally appropriate.
And the thing is, you actively are providing great rhetorical reasons to not listen to people who do have a sense of user empathy--which you lack, it seems, as you would rather browbeat people into using it than understand why they don't and maybe make it a better value proposition--and want people to use Linux. You're literally, precisely acting against the interests you purport to hold!
When you're in a hole, just...stop digging, dude.
Re-reading this discussion only shows one of us being unreasonable and abrasive. I don't care to engage you any further unless you change your attitude.
But it's me being unreasonable. And abusive--I laughed aloud, dude.
I have had performance decreases and increases with Wine. I would not call the change drastic. The larger issue with wine is that not every game works.
I encourage you to measure both and see what the performance looks like in both places, then see which one gives you better options for improving that if performance really matters.
Unless Linux has exclusives (which it doesn't afaik) then it has lost the battle completely and utterly. Windows is the default choice for gaming PCs and unless something drives a significant change that will continue to be the case as linux is clearly a second class citizen.
As a user, it depends on what you're looking for. I bet most end users would have no problems on a Linux desktop, especially one like Cinnamon or KDE Plasma.
I have no idea what you're talking about, Linux on the desktop is great.
The main problems with desktop Linux as I see it are 1) crappy DEs (namely Gnome3, the most popular and pushed hard by Red Hat) and fragmentation in the DE space, video driver problems, and power-management problems (particularly on laptops, esp. with suspend/resume). My laptop also has a lot of trouble going from docked (w/ dual monitors) to undocked and back. This stuff actually works pretty well on Windows, but on Linux it's a real PITA and frequently requires reboots IME, and that's on a not-new Dell business laptop that's generally well-supported.
I agree with most of your gripes, but fragmentation is more an more just becoming a reality. UIs won't look the same anymore, it just doesn't make sense to. So people will experiment with different things and try to find something better, the alternative is to stagnate with accepting what we have.
Fragmentation, can be restated as "Community diversity" and it ain't going away because its a huge strength. Because while you think gnome3 sucks if Canonical had their way everyone would be using it.
>Because while you think gnome3 sucks if Canonical had their way everyone would be using it.
Not true; if Canonical had their way, everyone would be using Unity. Honestly, I don't begrudge them making Unity; it's good to have some diversity, but I do begrudge them switching to Gnome3 recently and abandoning Unity altogether. A lot of people liked Unity and its concepts. Instead of throwing in the towel and becoming yet another me-too Gnome3 distro, they could have adopted KDE, and then created a custom KDE theme or version of Plasma which looks and works like Unity and allows them to continue to explore those concepts which leveraging the existing KDE underpinnings and the superior Qt toolkit. The real problem IMO is the horrible Gtk toolkit: it's in C but bolts on a hacky OOP concept, it's undocumented, and it's constantly changing. I know of at least one commercial project which had to abandon it and switch to Qt because of the constant API deprecation, and there's at least one FOSS DE (Lxde I think) which also switched from Gtk to Qt for the same reasons, and got a performance boost in the process IIRC.
>the alternative is to stagnate with accepting what we have.
I can't even count the number of comments I've read from people who are sick and tired of the constant change and churn in Linux UIs. Lots of people were perfectly happy with Gnome2, but then it was deprecated and they were pushed to Gnome3 and it was terrible and they hated it. As a result, not 1 but 2 new DEs were created, just to satisfy people who wanted something more stable: Cinnamon and MATE. It's a complete mess. A lot of people were similarly pissed when KDE4.0 was released and they couldn't use KDE3 any more because the distros all deprecated it, and the new version was filled with huge bugs and didn't work. At some point, you need to aim for product maturity instead of constant churn. I've been using ssh for many, many years and I don't see that being deprecated any time soon and having to switch to some other standard, but somehow we're supposed to accept this as normal for DEs?
3 months ago I would have agreed but they have announced that for their next LTS 18.04 they are dropping unity in favor of Gnome3.
I agree that improved interop would be good, but I am grateful for how well it does work. Copy/paste and other common shortcuts and features work. But, I use KDE and it sucks when I click something and it tries to open nautilus and clobbers my wallpaper. A "killall nautilus" later and everything is fine, but its annoying and a non-power user couldn't do this. The new the changed right click menu on the desktop would be really weird for the uninitiated too.
> I can't even count the number of comments I've read from people who are sick and tired of the constant change and churn in Linux UIs
It's not just Linux UIs Every UI is advancing; my friends and family are constantly asking me for help with win10 or whatever crazy mac/iOS/android shit they dig up. I explain I never used it, but I can form better google searches than them so I can figure out how to do what they need (I also don't mind this because I like learning this stuff too).
The root of this is how does a UI designer improve the UI while keeping it familiar enough for previous users? How do you keep power users and people on the cutting edge happy without losing the technophobic grandparents and those slow to learn?
No, it's still true. They didn't want to drop Unity at all; they did it because they were forced to. Didn't you read Mark Shuttleworth's blog postings about it? He was clearly unhappy about things. They gave up on Unity and Mir because they weren't working out in the market and it was costing them too much money to pursue that direction. If Mark had his way, Unity would be the most popular Linux DE.
>I use KDE and it sucks when I click something and it tries to open nautilus and clobbers my wallpaper.
Huh? I don't know what to say to this. Nautilus is Gnome software, not KDE. Why would KDE try to open Nautilus for anything, unless you've explicitly set it up that way? Now if that's what you've done because you like Nautilus, the problem is probably with the Gnome/Gtk devs. As I said before, they're infamous for not being amenable to cross-toolkit efforts. I would encourage you to ask them for their advice on this problem; they'll probably tell you "switch to Gnome3". :-/ I use KDE too, and personally I avoid GTK software if I can help it.
>A "killall nautilus" later and everything is fine, but its annoying and a non-power user couldn't do this.
I agree entirely. It's garbage, and it shouldn't be this way. The Gnome devs are entirely to blame for this mess, along with all the distros that push Gnome3. They don't even want you to use a wallpaper with Gnome3; Jon McCann said so some time ago, saying that customization (even wallpapers) is bad because it makes different Gnome installations different.
>It's not just Linux UIs Every UI is advancing; my friends and family are constantly asking me for help with win10 or whatever crazy mac/iOS/android shit they dig up.
Oh I entirely agree; all the UI people across the industry have gone collectively nuts. But I expect better from FOSS, for multiple reasons: 1) FOSS isn't for-profit (well, I guess Red Hat is, maybe that's part of the problem), so they shouldn't need to push unnecessary churn to justify pushing customers to "upgrade", and 2) FOSS is usually very resource-limited compared to commercial software, so unnecessary churn is a waste of developer time that's better spent fixing bugs, improving docs, making things more stable, etc. If I want to use the latest fancy BS, it would stand to reason that I'd go for commercial proprietary SW, because those vendors are all in favor of making "new" shiny stuff to get me to spend more money to do the same thing. The same shouldn't also be true of FOSS, but for some reason it seems to be.
>I also don't mind this because I like learning this stuff too
I don't. I consider it all an unnecessary waste, and I really don't like these new UIs at all. Thankfully I'm still stuck on Win7 at work for the moment, and minimize my use of Windows whenever possible as I do most of my work in a Linux VM (and don't use it at home at all), and I never even see any Macs up close. I put up with Android because it's the best of 2 bad choices, but also IMO it's not that bad: it's a touch UI designed for a handheld touchscreen device and it's OK at that, unlike PC stuff that they seem to be trying to tablet-ify.
>How do you keep power users and people on the cutting edge happy without losing the technophobic grandparents and those slow to learn?
Whatever it is, the answer isn't trying to tablet-ify PCs, or using garish and nauseating color schemes. Everyone was doing just fine with the WinXp-7 style UI for the most part, and why do we care about grandparents? Today's grandparents were using Win95 and XP, so they're perfectly used to that kind of UI. Grandparents today are not starting to use computers for the first time; that was 10-20 years ago. Anyone left now who isn't familiar with PCs to some extent isn't going to be alive for very long, and isn't interested in learning PCs now. And most older people with simple needs would be perfectly happy with an iPad or other tablet anyway, which is why the PC market is no longer growing.
It wasn't KDE software opening nautilus. It was random 3rd party software that decided to launch nautilus instead of asking what it should have launched.
> Everyone was doing just fine with the WinXp-7
Sounds like someone who never worked on a help desk. There is someone new using a computer for the first time everyday. I would be confident this was true even if it were just grandparents, but there are poor subsistence farmers settling into cities and luddites turned everyday. I think you are letting your tiny view (we all have a tiny view) of the world make you think that every hold a decent professional job in a decent western and industrial nation.
And no matter how much you long for it the past is not coming back and any UI that requires clicking "start" to shutdown is in wastebin and rightfully so.
Citation needed. Windows 10 still requires this, they just don't label the button "start" any more (nor does any other UI I've seen in the last 10 years). Mechanically, it's basically the same.
It still has a few dialog from the nt4 era, some that use the snap-in nonsense, regedit is still a thing. Then for about half of what a typical user does they can use the nice pretty UI, as long as there are no unexpected errors.
No, but it is the normal result of an open source development process. Being able to fork at any time is exactly why the fragmentation exists. Don't like it? Use a commercial product.
Consider how different the iPhone and windows 10 are. Or windows 7 and windows 8. Because the community driven teams don't stop when they run out of money they stick around, but they also have really small marketshare.
Honestly, this whole "WinDoze is lame" attitude is so 2008. Windows 10 is a great platform. And I say this as an Arch Linux user who just wasted 2 hours dicking around with DBus and NetworkManager because there are still 3-year-old unfixed bugs regarding Cisco VPN connections.
I really don't know why people like windows beyond having their applications (That really is enough). It is limited, the tools they make are clearly more for checking off boxed on some marketing list somewhere because they rarely work right and often limited in their core performance. Even their tools like their expensive compiler and database does work as well for my worklaods as their free and open source counterparts.
It wasn't that long ago we saw the remotely exploitable flaw in Windows Defender that allowed arbitrary execution of code in a security product! Or more innocuous but no less annoying, the guy with the umpteen core ryzen getting blocked by some random mutex deep inside windows so his build was being throttled and he couldn't use his own CPU the way he wanted. It buggy, its lower quality, but they succeeded in having many make their software for it exclusively so it is
Then we can start talking about the ethical issues with it, like giving money to a monopoly that has tried to bribe standardization bodies and has been convicted several times of anticompetitive practices.
I used Linux for over a decade exclusively while good friends told me how great OS x was now. I finally had to support building an iOS app so I bought a MacBook.
Great hardware. the os is awful compared to Linux. the window manager is worse.
they have some really nice apis available. increasingly to use those you must sell your software in their store and give them 33% of your revenue.
I think a world where one or two companies get 33% of all software revenues is roughly maximally dystopian.
On the other hand unless a linux package has a post-install script that moves things around, I can just dump where everything landed by a simple package query. To each his own, I suppose.
However I was comparing the newer package manager and Windows App Store install models to Linux's package management, since that is the most comparable model. Both installation models are closer to the Mac model, where an app package gets installed to a single location. Universal apps are containerized, and don't use the Registry at all[0]. The PS package manager installs packages into a single location, where you can pretty easily remove them using the package manager, or manually, if you want.
[0] http://www.tomsitpro.com/articles/windows-10-app-development...
This is news to me, unless its another one microsoft's crappy marketplaces, and I would like to do some reading.
https://docs.microsoft.com/en-us/nuget/tools/package-manager...
I much preferred the UI/UX of the old Classic Mac OS. Apple should have just plopped that on top of Darwin (I would probably still be a mac user).
Fedora is good enough, and definitely breaks less often than a Windows install (which is a big change from even a few years ago).
My experience is completely different. Not with Fedora, but with Linux in general. I've never actually managed to hose my Windows system. Closest I've gotten is when installing custom themes on Windows 7 and done fucked my explorer.exe, so my GUI was gone, but I was able to fix it with cmd. On other hand I've mauled multiple Linux installations to a point where screens are just black or login screen won't even show or system doesn't even recognise MBR on the disk.
OSX so far I haven't been able to break, but it might be just because I haven't actually tried to tweak it in many ways. Currently I'm kind a tied in the Apple ecosystem, so I just run OSX on my laptop/hobby machine and Windows on my "gaming rig", but I use Linux daily for work.
I do own a couple of gaming consoles, but now that I have less spare time, my gaming preferences have definitely shifted toward casual games I can pick up and play for 20 minutes and then forget about for a week (so I'm a Nintendo fan these days).
And the "this year is the year of the Linux desktop" (or gaming desktop) line has been going on since 1998. I was there, got the t-shirt.
We have a multiple GPU compute nodes with Teslas for research. The vast majority of us uses Macs.
Since I am only familiar with BMW attempts at JavaFX (described at JavaONE) and from what I am aware most embedded GUIs aren't X compatible, although there used to exist TinyX.
So is Qt kind of winning all the new GUI projects in that domain?
But do they use XWayland? Seems nvidia is still supporting X, Wayland, and Headless Compute modes of operation...
The more relevant question is "who wants a desktop Linux system". And the answer is very few.
That's why they doesn't care about that market. The huge majority of desktop users will go either Windows or OS X.
(And, of course, Nvidia is the only choice for now if you want GPGPU computing).
Desktop Linux with Wayland has maybe 0.1% market share rising to maybe 1% over the next years in the best case. In the worst case, Wayland will just be another Linux Desktop fail. Other devices (Phone/Tablets) do not use Wayland.
The amount of resources any given company can allocate to Linux platform bullshit is limited. The "serious" NVIDIA Linux customers will probably stick with X anyway, in the foreseeable future.
I'm not ruling out that there exist other pointless mobile operating systems with 0.0% market share that use Wayland or even XWayland in some capacity.
My point however is that the "market" argument is nonsense when the market is practically non-existent.
Don't turn a company's laziness into a virtue.
Didn't mean to do that. The whole thing has been a point of contention for a long time, so I wanted to clarify that they are actually getting on board with Wayland. They are by no means fully backing with Xwayland. I want a full Wayland experience and nVidia is doing the bare minimum to enable it on their hardware. For me getting all the way there really does involve desktop applications embracing it and adding support too. X is deprecated, and so should be Xwayland.
So: it could well be that this is an issue of resource allocation more than anything else. Over time, most applications should move to using wayland (or whatever window / surface library emerges as the clear winner, but that looks like wayland is the future here) for their graphics. A lot of effort into supporting Xwayland would mean work spent on a library that should be of limited utility - only useful during the transition between platforms (like the old carbon layer for OSX in the early 10.0, 10.1 days). Right? Or maybe I'm misreading it.
Xwayland enables these types of applications to continue working in a Wayland session, so that desktops and distributions can move to Wayland without having to either wait on the legacy applications or remove them from the distribution.
tl;dr Nvidia has no need to play in this space, anyone using X will just use X without the translation layer.
From that point of view, "Wayland" is kind of a short-hand for a number of different stacks that have some components in common, and speak the Wayland protocol. If you choose not to use a complete stack on a system, then the result can't be fully functional.
Personally, I don't have any sense of how much dark matter is in the area of Linux GUI applications. We can easily know about Open Source, where most modern applications sit on toolkits and desktops that are moving to Wayland, but presumably places like CERN have some really old internal applications that will only ever run on X until they are replaced. I'm genuinely unsure how large that problem is.
I have to say that I'm always skeptical about arguments that something can't be done for performance reasons. People have been using that argument against new technology since at least the days when C was decried as new-fangled, decadent luxury.
There is no standard way to register global keybindings under Wayland yet, but this is bound to happen eventually. Not as a builtin part of Wayland, but it doesn't matter. It will of course use D-Bus interface, as KDE does it now for example - but if an application doesn't support this interface, it can't steal keybindings, it just doesn't get to install global keybindings itself.
Not sure what you have exactly against D-Bus, but clearly modern Linux desktop embraced it.
In the desktop things, that wayland aims to do, dbus is useless. (actually i don't know what it's useful for at all, other then quickly hacking some object oriented (gui)client-server desktop program)
I do recommend reading the dbus protocol (the protocol, not the API).
Sorry to sound negative but "modern linux" now means "desktop crap", and it is all crap. Freedesktop.org (the linux "desktop" "standards" authority) has gone from "ehh, passable" to "complete crap" in the last... idk 5-10 years (the "desktop" quoting is because it has gone pass the actual desktop problems).
Qt and GTK are not responsible for implementing the compositor at all. Compositors aren't actually the problem at all, the problem is applications.
XWayland is one of the main ways we are going to bring legacy applications to Wayland desktops. It works about as well as X does, and lets the other applications take advantage of the sane architecture of Wayland.
> In my view desktop/workstation use cases are not sufficiently covered by Wayland, and that is where nvidia's focus has been.
What use cases? The only thing that's missing for me is input method integration. It has full display colour management, handles multiple refresh rates in the same session well, handles scaling, supports every input device I own (including graphics tablets). What's not really firm right now is native support in every application, that's coming along at some rate, and XWayland is bridging that gap quite well.
Probably the most salient issue with wayland is that it prevents programs from communicating with each other in the name of 'security' and 'sandboxing' when preventing those things at the display server level adds literally zero security and just makes everything more painful for application developers (imo to the point where they simply won't touch the stuff). If there were a 'windowing protocol' that went along with wayland that covered many of the things wayland does not then I do not think there would be as much of an issue, but the thought that we can force a new display server on people without the other bits as well seems at best naive. Again, I think that the absence of such a 'windowing protocol' from wayland is because one of the main use cases it was developed for often had literally one program running (car backup camera displays).
I do not dispute that xwayland bridges the gap. But if people are using X right now then nvidia has no reason to support xwayland right now because there are no wayland only applications that their customers are demanding support for. Down the line that may change, but from a corporate decision making point of view there is zero reason to do this.
You don't need to. Just like X11 has Xlib and libxcb, Wayland has libwayland. It is actually easier to get input from libwayland than it is to get it from Xlib, and about on par with libxcb.
> I realize this is a bit of an exaggeration, but X11 is more than something that draws windows and from my understanding wayland doesn't address many of the other parts. For example, from my understanding wayland does not provide a standard way to support window manager level keybinds that override the binds set by the developer of the window that currently has focus.
You're right about that, global keybindings will have to be allowed through the compositor, and I don't know if there's a protocol for that yet. Global keylogging is one of the things Wayland was designed to avoid, so the way we implement that feature will have to be a bit different. It's for good reasons though.
> For me this means that the primary way I interact with my computer is now broken on wayland unless every single program I use implements the ability to pass those custom keybinds to the program that actually handles them. Sure, everyone can implement their own way of dealing with window manager level keybinds but that completely defeats the purpose of having a standard in the first place.
I don't know that it defeats the purpose of a standard! I agree that these use cases need to be addressed, but I don't think the way to do that is to give every Wayland client the ability to globally change the way input reaches other clients without permission from the compositor.
> Probably the most salient issue with wayland is that it prevents programs from communicating with each other in the name of 'security' and 'sandboxing' when preventing those things at the display server level adds literally zero security and just makes everything more painful for application developers (imo to the point where they simply won't touch the stuff).
Yeah, I think that the XDG compositor extensions should loosen this stuff up a bit, or at least make it easy for an application to ask the user for permission to spoof everything. I'm unlikely to be displaying a hostile application on my machine. That said, doesn't exactly add "zero security" though. It is helpful for partially-isolated applications, like those in containers. I would like to be able to trust the isolation of applications some day so that I can run spotify on my workstation without breaking into a sweat.
I think it would be perfectly adequate to have the application ask for permission to do the heinous boundary violations that make X11 fun and dynamic.
> I do not dispute that xwayland bridges the gap. But if people are using X right now then nvidia has no reason to support xwayland right now because there are no wayland only applications that their customers are demanding support for. Down the line that may change, but from a corporate decision making point of view there is zero reason to do this.
For what it's worth, NVIDIA is looking to support Wayland on desktop. They already support it commercially in their embedded products (especially targeted toward automotive). So it's not exactly accurate to say that they have no customers asking for Wayland-only application support. The question is, when will it be convenient on desktop Linux? and that's hard to say. It is somewhat surprising that they aren't currently planning to support acceleration in XWayland, given that they have at least put some effort into supporting Wayland in general.
Basically, if you're not a gamer, the open-source Nouveau driver should provide all you need, assuming it works for your card. But parts of your system may stop working, or cause other bugs, from not using the official drivers.
edit
I might have misinterpreted your question, so here's an alternate answer. If you mean "Why can't we just do 3D rendering with a framebuffer?", a framebuffer is just a copy of the rendered screen in memory. You render it in your local CPU, and then copy it over into video card memory. So it doesn't use hardware rendering, and can work with pretty much any card.
If you do want to do hardware rendering (or most any feature of your card at all) you have to write a kernel driver (the DRM) to use the chip, and then some userland software (the DRI) to manipulate that driver, and send data to it somehow from your client app using a standard API for 3d graphics (OpenGL's GLX extension for the X11 protocol). The devil is in the details though, so to work around some limitations, some people wrote new graphics software to allow the client app to render without a "middle man" in the way. That's where you get Wayland.
Performance is good for regular desktop use, including video playback (then again, my CPU is a Ryzen 1700), 3d gaming performance is not overwhelming, though. Plus, Kerbal Space Program gives me weird display errors and sometimes manages to crash Xorg.
I am not much of a gamer, though, so I can live with this arrangement.
There's an effort to reverse engineer this called nouveau, they've done a surprisingly good job in a surprisingly short time, but it's still far from perfect.
Nvidia has been putting a lot of hurdles on their efforts since the Pascal GPUs. If you look and pre-Pascal GPU the performance of Nouveau is really impressive, but Pascal GPUs are crawling to go beyond a few frames per second. It's really sad.
But Mesa really caught up lately, so switching to AMD was never as good. Just forget Nvidia and enjoy properly integrated graphics stack.
It's not as if "we have no plans for support" could result in accidentally having support.
Reverse those numbers and you'll hear a different story.
"Speaking of the Linuxes, Ubuntu is tops among them with 12.3% of the entire OS market for developers. Fedora, Mint, and Debian accounted for 1.4%, 1.7%, and 1.9% of all responses, respectively."
No they don't. I have had 3 different nVIdia configs, and presently 2 for gaming, and through my experience hopping from one distro to the next I find that the problem is actually unrelated to nVIdia drivers - but the compositor. For example, in Mint if you switch to compton, it makes all tearing disappear. This is well documented.
If/when Wayland becomes the default for Ubuntu, Nvidia will follow. Can't let them ponder the amount of time it would take to code the missing layers for AMD hardware
I sometimes fantasize about hooking up twelve monitors for a wicked game of descent, though.
You're right that we don't use our computer card for our graphics and, as many people pointed out, the computation will wind up running on a headless server. On the other hand, during development, our code will be running on our desktops and will be using the GPUs on the graphics card. I've witnessed a developer's frustration with X11 driver issues cause a project to transition from CUDA to OpenCL.
Granted, he didn't just throw a hissy fit an port the code. Rather, it inspired him to write up a tradeoff analysis of OpenCL versus CUDA that showed that OpenCL offered this specific project significant benefits and that the main selling points of CUDA weren't being used in this instance. He also wound up rage coding a prototype port to OpenCL to go along with the proposal.
The project would have moved eventually, since it was the better engineering decision in this case, but it was about a half decade away on the timeline until the lead dev's compositor stopped work.
As a Fedora user I can understand the reluctance. The recent upgrade to Fedora 26 has given me a ton of gnome crashes which I think are Wayland related. Commentary on Fedora channels tends to be along the lines of "Wayland is great and works for everyone (OK, maybe not those with NVidia cards)".
If Ubuntu has millions of users it may have millions of NVidia users. Putting them onto unstable technology to pressure Nvidia into supporting Wayland wouldn't be a great move.
Therefore I can understand Nvidia's lack of haste. They will support the version/fork that Ubuntu chooses. No point in wasting resources until then.
Of course that would not happen if they opensourced their drivers but I kinda lost hope about that...
Ubuntu is the default distro for gaming, too, and I doubt they will jump until it's well ready, and certainly NOT before nVIdia provides support. It would make no sense.
* The big exception is Chrome which uses GLX for accelerated rendering. The software fallback path uses a lot more CPU and won't support WebGL AFAIK.
They continue to expand their team, and there is a certain amount of broader community involvement in the Mesa-based driver. Valve software has at least one full time developer working on the driver in addition to the full time developers working directly for AMD on it.
The peak performance of AMD's top (pre-vega) cards right now is less than the peak performance of NVIDIA's top cards, but for similarly priced cards, the performance is basically on par. Finishing out 2017, and starting 2018, I think AMD may reach or exceed NVIDIA in flagship performance.
This is true and I'm a happy linux gamer these days but I wanted to point out that these are new drivers and the mesa support is really only in new kernels. Ubuntu 16.04 - don't bother, Ubuntu 16.10 - getting there, Ubuntu 17.04 - worked great for me since day 1. Arch will also work fine.
I do play some games, but it's on a 3 year old desktop which is also running 4 fairly busy VMs (i7-4770 CPU @ 3.40GHz with 32 GB RAM and the built-in Intel graphics card).
Sapphire is one of the best card makers with AMD chips: http://sapphirenitro.sapphiretech.com/en/580-8.html
The only problem - lately they are hard to buy. I guess all those bitcoin miners buy all of them at once.
Laptops and media PC are all Intel and they are reliable as well.
Will depend on your use case though.
Nvidia supports X11, on Centos/fedora. ITs first class support and the driver is as fast as windows (or faster)
As of recently, I have a mixed 1920x1200 and 4K setup on my home desktop, and 1920x1200 + 2560x1440 setup on my work laptop. I was very interested in what I'd heard about Wayland supporting per-monitor scaling and hoping to move over whenever KDE's Wayland support stabilised.
Well, I guess I'll still be doing that on my work laptop with its Intel graphics... here's hoping AMD does better for my next desktop GPU upgrade.
And Wayland is a huge improvement over X. Whenever I go back to X, it really suffers from tearing on portrait sub-1080p DIV-D monitor.
The graphics market is dangerously locked up by two players and its time some new players took the capital risk and stepped in. Nvidia would probably give a shit if someone bothered to challenge their share of the market in both graphics and computational space.
I'm really glad I ditched Nvidia and switched to AMD. With Mesa progressing at fast pace, Nvidia will become simply inferior in many aspects, including the above one.
Maybe AMD if they drop proprietary drivers. But never Nvidia.